Embed Size (px)
Citation preview
Working Paper
Independent Spending in State Elections:
Vertically Networked Political Parties Have Been the Real Story
Keith E. Hamm Edwards Professor in American Government
Rice University Department of Political Science MS-24
P.O. Box 1892 Houston, TX 77251-1892
(713) 348-5273 [email protected]
Michael J. Malbin Executive Director
The Campaign Finance Institute 1425 K Street NW, Suite 350
Washington, DC 20005 (202) 969-8890
Professor of Political Science University at Albany, SUNY
Albany, NY 12222 [email protected]
Jaclyn J. Kettler Ph.D. Candidate Rice University
Department of Political Science MS-24 P.O. Box 1892
Houston, TX 77251-1892
Brendan Glavin
Data & Systems Manager The Campaign Finance Institute
1425 K Street NW, Suite 350 Washington, DC 20005
(202) 969-8890
March 2014
Revised version of a paper prepared for presentation at the 2012 meeting of the American
Political Science Association, New Orleans, LA, August 30–September 2.
Authors who choose to quote or cite the information in this draft working paper should be
aware that subsequent drafts may alter the data or text.
1
ABSTRACT: This article examines independent spending in state elections before and after
the Supreme Court’s 2010 decision in Citizens United v. FEC. We find that the decision did
not have much of a direct effect on business spending, despite public expectations. Increases
were higher in the aggregate in states that prohibited corporate spending before the decision.
However, the major growth was not in the business or labor sectors, but in political party
network organizations -- and most particularly the national organizations of state elected and
party officials. Contrary to some contemporary views about political parties, these
developments cannot be understood as a displacement of within-state money from parties to
interest groups. Instead, national party organizations were operating across state lines,
deciding whether to contribute to formal party committees or to their functionally-equivalent
party allies as local circumstances might dictate. The complex movement of money belies
any theorizing that would treat a decline in the proportional role of formal party spending as
equivalent to a zero-sum increase in non-party interest group power. Rather, we see the
pattern of independent spending as part of a larger story of change in American political
parties. These changes now include vertically networked political parties operating across
levels of jurisdiction, alongside the horizontal networks receiving attention in recent
scholarship.
KEY WORDS: campaign finance, Citizens United, business politics, interest groups,
political party networks.
2
There can be no denying the importance of what has been happening to campaign
finance in the United States since 2010. However, there is confusion about whether the
Supreme Court’s decision in Citizens United v. Federal Election Commission should be seen
as a primary explanation for what has occurred. This article begins to disentangle the strands
through an analysis of independent expenditures in elections at the state level in 2006 and
2010 from new data gathered and supplied by the National Institute on Money in State
Politics.
The paper’s findings tend not to support claims being made on opposite sides of the
contemporary political debate with respect to business and political parties. First, contrary to
statements made by some of the decision’s critics among reformers, the data presented in this
article show that Citizens United itself did not have a noticeable direct effect on business or
labor independent spending in 2010, even though the key holding of the case was about
corporate spending. Given that ruling, one might have expected spending to increase more
rapidly in states that previously prohibited corporate spending than in ones that always
allowed it. But it turns out that the increases in business spending were more or less
comparable across the two sets of states. This is consistent with past political science
research about business political activity, which has found most publicly traded corporations
to be institutionally cautious in their political activities, pursuing access-seeking goals rather
than more risky electoral ends. (Eismeier and Pollock, 1988; Sorauf, 1988; Boathright et al,
2003; Boatright et al., 2006; Clark-Muntean, 2008; Clark-Muntean, 2011; Issacharoff and
3
Peterman, 2013.) Given this past behavior, we should not expect most publicly held
corporations to change their behavior in candidate elections just because courts said that they
were now allowed to do so. We expect them not to increase their spending dramatically
because it is generally not in their interest to do so.
The paper also addresses a claim normally associated with quite a different position
in the contemporary debate. It is sometimes said that limiting contributions to the political
parties (ending party “soft money”) has displaced funding that would otherwise have gone to
the parties, redirecting it into spending by unaccountable non-party organizations (La Raja,
2013a and 2013b). We consider this to be too stylized a theory of the role and activities of
political parties today. Contribution limits may indeed have affected whether formal state
party committees or other all-but-formal party allies were the main independent spenders in a
particular state. But to think of this as a zero-sum trade-off enters the process at its mid-
point. Rather than donors within a state diverting their money, much of the increase came
from national party-related organizations that raised their money nationally and then decided
opportunistically which state races to enter. Once a race was selected, the organization
would then consider whether local circumstances favored sending the money through the
formal state parties or through non-party organizations that were closely allied with the
parties. Either way, the key actors were national.
This suggests a need to revise our thinking about political parties. In recent years,
scholars have written about the parties not simply as formal organizations but as overlapping
networks of participants in the broader political environment (Monroe 2001; Cohen, et al.
2008; Koger, Masket and Noel 2009; Masket 2009; La Raja 2013a; Skinner, Masket and
Dulio, 2013.) For the most part, these authors are writing about the horizontal networking of
actors within a single election or jurisdiction. The independent spending of 2006-2010
4
suggests something different and – vertical networking by actors who are operating across
state lines and across levels of jurisdictions, with their strategic and financial decisions made
nationally.
Background
Citizens United was one of two court decisions of early 2010 with a direct bearing on
this study. In it, the 5-4 Supreme Court held that corporations (and by extension labor
unions) had a right to use their own treasury funds to make expenditures expressly
advocating the election or defeat of a candidate, as long as the spending was done
independently of the candidate in question. Soon after this decision many commentators
speculated that it would result in a substantial increase in spending by large, for-profit
business corporations. This is one of the assertions we plan to test.
Citizens United applied not only to business corporations but to nonprofit advocacy
organizations. However, having the right to spend money does not do much unless one has
money to spend. Three months after Citizens United, a lower court made it much easier for
advocacy organizations to raise money for politics. In March, the U.S. District Court for the
District of Columbia ruled en banc in SpeechNow.org v. Federal Election Commission that
individuals have a right not only to spend unlimited amounts on their own, but to give
unlimited amounts to an organization to make those expenditures. Although the case was
brought before the Citizens United decision on behalf of individual donors, the decision’s
timing let the lower court rely on Citizens United, thus indicating that corporations and labor
unions could also make unlimited contributions to independent spending organizations. We
expected many of the beneficiaries to be ideological or issue groups, like the Speech Now and
Citizens United organizations themselves.
5
Data and Methods: This paper uses state elections to disentangle the growth in
independent spending that can be attributed specifically to Citizens United. This would not be
easy to do through federal elections, but state elections afford us leverage in approaching the
issue. Twenty-five states had bans or limits on corporate and/or union independent
expenditures at the time of the Citizens United decision. The Court’s elimination of these
bans provides a quasi-experimental research opportunity to investigate the impact of altering
the rules. If Citizens United had a strong and direct impact on the level of independent
spending, then the twenty-five states with a ban in place before the decision should show a
higher rate of spending increase than the twenty-five states without such a ban. These
twenty-five states with a prior ban should also show a higher rate of spending increase
among business organizations than among others. But if independent spending increased in
the states, and the presence or absence of a previous ban does not help explain the increase,
we will have to look for other explanations. The most likely source of growth would be
among non-business organizations that were able after SpeechNow to accept unlimited
contributions. We do find this, but we also find that the steepest growth occurred not among
the issue and ideological groups predicted but among organizations closely related to the
major political party networks.
Unfortunately it is not possible to test these thoughts by comparing all states. Not
every state requires disclosure of express advocacy independent expenditures and fewer
require it for what federal law would call electioneering. Moreover, the disclosure records
across states until recently had not been maintained in a common format. After Citizens
United, the National Institute on Money in State Politics (www.followthemoney.org) decided
to collect 2006-2010 independent spending information for twenty states and to make the
data available to scholars. The laws of these states included a mixture of disclosure practices.
6
Some required disclosure for independent expenditures but not electioneering. Others
changed their disclosure laws between 2006 and 2010. Given the variations in states’
definitions of the terms independent expenditures and electioneering communication, we
refer to all such expenditures as independent spending (IS) but we only compare similar
forms of spending in any given state across the years. For general comparisons, that permits
us to consider independent expenditures for sixteen states1. Electioneering is only included
for states with similar data across election cycles.2 We also omitted 2008 for this article.
That lets us compare elections before and after Citizens United (2006 and 2010) when all of
the states among our sixteen were at the same point of their gubernatorial cycles. The second
part of the paper, on the offices and candidates being targeted, includes fifteen states instead
of sixteen because the disclosure documents for Florida contain no information about the
races toward which expenditures were directed.
Initial Analysis: Before and After Citizen United
The first question is whether independent spending in state elections increased after
the Citizen United decision, as is commonly stated. The answer in the aggregate is clearly
yes. In the 16 states with comparable information for the two years, independent spending
rose by $62 million over the four-year period, an increase of 43%. By contrast, contributions
to candidates over this period in these same states increased by only 14%.
The timing straddles Citizens United, but does the decision explain the increase? To
answer this, we divided the states into two groups: ones prohibiting corporate spending and
1 Alaska, Arizona, California, Florida, Iowa, Idaho, Maine, Michigan, Minnesota, Missouri,
North Carolina, Ohio, Oklahoma, Tennessee, Texas, Washington.
2 California, Florida, Iowa, Ohio, Oklahoma, North Carolina, Washington.
7
ones allowing it prior to the decision. If Citizens United were behind the increase, we should
see it primarily in the ten states that prohibited the spending before Citizens United.3 The
results shown in Figure 1 indicate that IS increased significantly in both groups of states. In
ones where corporate spending was prohibited in 2006, the increase from 2006-2010 was $24
million or 65%. But spending in the six states4 where corporations were permitted to make IS
before 2010 also increased $38 million or 36%. Thus, while total spending by all sectors did
increase more in the aggregate in states with a prohibition, the fact that a significant increase
occurred across both sets suggests we should look in more detail.
[Figure 1 goes here.]
Independent Spending by Sector
Any useful explanation should begin by disaggregating the independent spenders.
We decided to focus on business and labor groups (the ostensible winners from the CU
decision), ideological /issue groups, and three sectors related to the political parties.
Clarification of the sectors is provided below:
– “Party” refers to official state and local party committees.
– “Party-affiliated” refer to national political committees such as the Republican
Governors’ Association (RGA), Democratic Governors’ Association (DGA),
Republican State Leadership Committee (RSLC) and other entities clearly affiliated
with one or the other major party. (The RGA and DGA once were formally parts of
the Republican and Democratic National Committees but separated after the
Bipartisan Campaign Reform Act of 2002 prohibited these committees from raising
soft money for non-federal purposes.) 3 Alaska, Arizona, Iowa, Michigan, Minnesota, North Carolina, Ohio, Oklahoma, Tennessee, Texas.
4 California, Florida, Idaho, Maine, Missouri, Washington.
8
– “Party-allied” refers to non-party organizations pursuing broad candidate and issue
agendas but whose activities and missions make them strongly associated with one or
the other party, rather than an issue arena. Organizations in this category may not
support all of a party’s nominees, but they invariably support nominees from only one
of the parties. An example on the national level would be American Crossroads,
which supports Republican candidates for many offices.
– “Ideological/Issue” refers to organizations strongly associated with ideological or
issue positions. These organizations (whether “progressive” or “conservative”) may
support candidates from one of the parties most or even all of the time, but their
identities and goals are distinct.
– “Business” refers to General Business groups, such as the Chamber of Commerce, or
sector groups – whether in Construction, Finance, Insurance & Real Estate, Health, or
other fields. (Individual corporations tend not to make independent expenditures.)
– “Labor” refers to individual labor unions or coalitions of unions, if the information
about the labor affiliation is provided in the group’s name or description.
[Figure 2 goes here]
The results shown in Figure 2 are unanticipated. Independent spending across the 16
states rose for four sectors and declined for two. Independent spending by business
organizations across all 16 states actually decreased from 2006-2010. The figure subdivides
the labor and business bars to show spending in states with and without pre-Citizens United
bans on corporate or labor spending. The business bars show that spending by business
organizations stayed level across the ten states that previously prohibited corporate
expenditures, while it declined in the six states that permitted the spending before Citizens
United. (The steepest declines were in Florida and Ohio. See appendix figures A-1 and A-2
9
for state-by-state details.) If there was any increase in business spending, it was not
happening through business organizations.
Labor spending, by contrast, increased dramatically from 2006 to 2010. The increase,
however, did not come in those states which had previously banned labor independent
spending. Rather, the increase was almost entirely due to California, a state where labor has
been active in independent spending for years. Moreover, a significant portion of the shift in
California came from a shift in labor’s strategy: in 2006 unions gave millions to two
organizations we categorized as ideological / party-affiliated (Alliance for a Better California
and Working Californians) while in 2010 labor’s IS money in the state went into direct
expenditures by the unions. In other words, labor did not so much change their level of
activity as change the vehicle. If the data for California are removed, labor increased from
$3.7 million in 2006 to $6.9 million in 2010, with the increase approximately equal across
states that previously prohibited labor spending or allowed it.
Political Parties and Their Allies
The major sources of increased independent spending in our states were not the
business, labor, ideological or issue groups. Rather, they were among groups we might think
of as parts of extended party networks – the formal political party committees and what we
call the party-affiliated groups (such as the RGA and DGA) and party-allied groups (such as
Crossroads). The increases were sizable for parties and party allies. They were even larger
for the party-affiliated groups, such as the Republican and Democratic Governors
Associations. (We have more to say about these below.) But the various types of party,
party-affiliated and party-allied groups did not behave the same way in all states.
As a working hypothesis, we expected to see political party contribution limits and
independent expenditure patterns relate strongly to each other. We expected parties to make
10
more independent expenditures in their own name where they can accept unlimited
contributions. In states with party contribution limits, we should see more spending by
groups allied with but separate from the formal party organizations. In fact, we expected not
only to see more independent spending by non-party groups in the party-limited states, but
we also expected a greater increase in independent spending by party allies in the party-
limited states after Citizens United and SpeechNow because the allied groups should find it
easier to raise money after the two decisions.
While the expectations are plausible, the results do not show the expected
relationship. Figure 3 shows independent spending by formal party organizations, party-
affiliated groups and party-allied groups in 2006 and 2010. The three pairs of bars on the left
show the results for the four of our sixteen states whose laws put limits on contributions to
the party. The pairs on the right show the results for the twelve states with no such
contribution limits. Because the states and their parties are of different sizes, and the bars
show the cumulative totals for all of the states, the left and right hand set of states are not
directly comparable to each other. However, was can learn something from the relative
heights of the bars within each of the subsets across the two elections.
[Figure 3 goes here]
The results shown in Figure 3 are surprising. Our assumption led us to expect
proportionally more independent spending activity by party-affiliated and party-allied groups
in states where contributions to the formal parties are limited. Instead, we see very little
spending by party-allied groups in either year in these states and little party-affiliated group
spending in 2006. Surprisingly, we see proportionally more party-affiliated and party-allied
spending in 2006 in the states without party limits than in states with them. This throws a
monkey wrench into the notion that political party contribution limits are the key mechanism
11
driving money away from the formal party committees. Independent spending also went up
in 2010 for both party-allied and party-affiliated groups in the states without party limits.
The one category that seems to follow our expectations is that of the party-affiliated
groups in 2010. But before we start treating one example as a universal law, we have to note
how much of the increase is due to a single organization, the Republican Governors
Association (RGA). Consider these facts:
• In 2006, party-affiliated organizations made $208,099 in IS in states with limits on
parties. In 2010 the figure was $17,106,913.
• The RGA made no independent expenditures in these four states in 2006 (and only
$244,528 in states without party limits). In 2010 the RGA spent $12.8 million in the
four states with limits and nothing in the states without limits.
• The RGA’s $12.8 million made up three-quarters of the total IS by party-affiliated
groups in the four states with limits. (Interestingly, the DGA had no direct IS in our
16 states, preferring to give to progressive coalition groups.)
The presence or absence of party contribution limits undoubtedly had something to do
with the RGA’s spending decisions, but the RGA was not the only party-affiliated
organization with increased IS in 2010. The Republican State Leadership Committee – an
organization focused on legislative races – spent $205,000 in 2006 compared to $2 million in
2010. But in contrast with the RGA, $1.8 million (90%) of the RSLC’s spending in our 16
states in 2010 was in states without party limits. So, as we noticed before when we compared
states with and without corporate spending limits before Citizens United, something more
complicated is going on.
Diverse Spending Paths: One problem with empirical analysis and interpretation stems
from the fact that the labels in disclosure documents are often inadequate to tell us what is
12
happening. For example, the Republican Governors Association spent a combined $6.6
million through two other organizations in the state of Ohio to influence the very close
contest between the successful GOP candidate John Kasich, who won with 49% of the vote
(and who raised $18.3 million for his own campaign), and the unsuccessful Democratic
incumbent, Ted Strickland, who received 47% (and raised $16.6 million). The money is not
reported as independent spending by the RGA but by the organization to which the RGA
gave money. Supporting the Democratic Strickland was an organization that existed for only
one election cycle called Building a Stronger Ohio. Building a Stronger Ohio had receipts
and expenditures of $4 million, with $3.8 million in IS. Of the organization’s $4 million in
receipts, $3 million came from the Democratic Governors Association and another $785,000
from labor unions. Thus, an organization can choose to support independent spending
through direct spending or through contributions to others who make those expenditures.
In Ohio, the RGA and DGA chose to give their money to allied organizations rather
than contribute it to the state party or spend it directly themselves. Their money followed
similarly layered paths in Minnesota, Michigan, and other states. But the RGA’s and DGA’s
strategies were not the same as each other within many states, and their decisions varied
across states.
[Figure 4 goes here]
Figure 4 shows that the RGA put most of its money nationally into political
consulting or media. (Independent spending would fall within this category, but so would
issue advertising and other direct public communications.) Its second highest level of
spending was for direct contributions to political parties. The DGA, in contrast, channeled
most of its money through other political groups. The DGA’s strategy had little to do with
political party contribution limits, since it pursued the same strategy almost everywhere. The
13
RGA clearly was giving its money to the parties as contributions only where the law
permitted it to do so, but it was also spending money directly and through other intermediary
organizations in the same states. It would be a mistake to say that state contribution limits
were diverting the RGA’s donors away from the formal state parties and toward the state-
level independent spending committees. That misses a key step. Donors gave unlimited
contributions to the RGA and DGA, both of which are national committees organized to
influence state elections. The donors clearly expected the organizations to behave
strategically to affect gubernatorial elections in whatever way made the most sense
politically. The donors were not steering their contributions depending upon state law. They
were giving to a national organization (the RGA or DGA) which in turn decided the most
effective way to spend it. This is a perfect example of the concept of vertical networking.
Earlier we showed that increases in independent spending by business organizations
did not track the states whose laws were changed by Citizens United. Since the RGA and
DGA were so important to the aggregate increase between 2006 and 2010, one rejoinder to
our claim about business spending might be that business in effect laundered its money by
funneling it through the most active organizations in state elections. Table 1 shows the
sources of contributions to the RGA and DGA in 2006 and 2010. The table shows that the
major donors to both the RGA and DGA were a mixture of individuals (with most of the
money coming in large contributions of $75,000 or more), corporations, and (for the DGA)
labor unions. The proportions among these types of donors were roughly the same in the
2006 and 2010 elections. Between 2006 and 2010 both organizations doubled their corporate
intake, but the RGA also more than quadrupled its receipts from individuals. Some of the
RGA’s increase came from the top individual donors. The late Houston home builder Bob
Perry, a frequent donor to Republican and conservative causes, topped the list in both years
14
with gifts of $2,050,000 in 2006 and $8 million in 2010. In our view, it would not be
accurate to interpret Perry’s contributions as if they were corporate contributions focused on
real estate interests. Like other wealthy individuals who own and run privately-held
corporations, Perry’s contributions were more partisan and ideological, In addition to the old
mainstays like Perry, the RGA under the leadership of Mississippi Gov. Haley Barbour also
persuaded new donors to give. In 2006, the top 20 individual donors gave $8.8 of the $9.9
million from individuals. In 2010 the top 20 individual donors gave $20.8 million of a total
of $42.8 million from individuals. (For a law review article about the RGA and DGA from
2002-2010, see Torres-Spelliscy, 2012.)
[Table 1 goes here]
The RGA did also have some important business contributions, but interpreting these
is not straightforward. The largest organizational business contribution to the RGA in 2010
was from the Michigan Chamber of Commerce for $5.4 million – more than four times as
large as the next highest. Yet when we look through our database of disclosed independent
spending, we see no activity in Michigan by the national RGA. Uncovering what happened
took some doing. According to an investigative report by the Michigan Campaign Finance
Network (MCFN), the RGA’s Michigan PAC (which does not accept corporate money and is
legally distinct from the multi-state “527” political organization5 discussed so far) received
98% of its $8.4 million in income from major out-of-state individual donors (including the
previously mentioned Texan, Bob Perry, and another familiar conservative donor, David
Koch). The RGA Michigan PAC then gave $5.2 million to the Michigan Republican Party
(which is allowed to accept unlimited contributions). The state party also happens to show up
5 For a definition of type of organization commonly referred to as a “527” committee, see U.S. Internal Revenue Service, “Definition of a Qualified State or Local Political Organization,” at http://www.irs.gov/faqs/index.html. Accessed August 25, 2012.
15
in our data with just about the same amount ($5.2 million) in independent expenditures. The
Michigan Republican Party also gave $3 million to Texas Gov. Rick Perry’s 2010 reelection
campaign. Finally, the RGA’s national 527 committee, (the one with the $5.4 million
contribution from the Michigan Chamber of Commerce) spent at least $3.5 million in issue
ads in Michigan that were simply not disclosed through the state’s campaign finance reports
because they did not count legally as political advertising (Michigan Campaign Finance
Network, 2011). In other words, at least some portion of the contributions being given to
help the Michigan party was actually Texas money being rerouted back into Texas, while the
Chamber of Commerce’s spending was directed toward spending that would not be reported
in any database as being political. A 2014 New York Times article contained additional
examples of nationally directed contributions being funneled through layers of state
organizations, indicating vertical as well as horizontal networking (Confessore, 2014).
Working with the disclosed data therefore can be suggestive, but the data should not be
treated as if they tell us all we need to know to trace vertical networking or compare sectors
across time.
Which Offices Were Targeted by Whom?
We focused on the RGA and DGA in the previous section because of the dominant
role they played in the increased spending in our states between 2006 and 2010. As is evident
from their names, the two organizations were most active in competitive gubernatorial
elections. Gubernatorial and legislative elections together accounted for 88% and 87% of all
independent spending in our states in 2006 and 2010, with roughly similar amounts spent on
the two branches. We therefore wondered what factors explain the decisions by independent
16
spenders to become involved in a gubernatorial election, and whether the considerations are
different for legislative elections.
Gubernatorial Elections
Our examination of party spending in gubernatorial races considers the effects of
electoral and institutional variables on independent spending by party and party-network
groups6. We conduct a multivariate analysis with independent spending by party and party-
network groups in gubernatorial races as our dependent variable7. To make the data more
comparable across states, we divide spending by the total number of votes in the
gubernatorial election. The mean of this dependent variable is $3.76 per vote with a range of
$16.80 per vote, ensuring we have plenty of variation to study despite the small sample size.
We hypothesize that the electoral and political environment will affect the level of
independent spending by party and party-network groups in gubernatorial elections. Our first
electoral independent variable of interest is the total amount of candidate contributions per
vote. We expect that the more money the gubernatorial candidates raise, the more likely it
will be that independent spending by party and party-network groups will also be higher.
We also hypothesize that the level of competition in a gubernatorial election will have
a significant impact on the willingness of party and party-network groups to engage in
independent spending. We measure the level of competition by calculating the margin of
victory between the top two vote-getters. We expect the margin of victory to have a negative
relationship with party and party-network groups independent spending.
6 For the remainder of the paper, when we refer collectively to independent spending by political party committees, party-affiliated groups, and party-allies we use the phrase “party and party-network groups”. 7 Some of the 2010 Republican Governors Association spending in Ohio did not include target information. We assume this $3.3 million was spent in the gubernatorial race, because the rest of their $8.4 million independent expenditures were spent in the gubernatorial race.
17
Our multivariate analysis also includes two institutional variables that may affect the
amount of independent spending by party and party-network groups in gubernatorial
elections.8 The first is control over redistricting, which we include in an attempt to measure
the relative value of the governor’s office across the states to the organizations doing the
spending. Our redistricting control variable sorts the states into three categories: political
control over redistricting, a hybrid system, or states with a redistricting commission.9 Party
and party-network groups should be interested in control of the redistricting process because
it affects the party’s ability to gain or maintain a majority in the legislature. Thus, we predict
that party and party-network groups will do more independent spending in states with
political control over redistricting.
The final independent variable of interest is the presence or absence of limitations in a
state’s campaign finance laws on contributions to political party committees. Where
contributions to the parties are limited organizations within the party network other than the
formal party should be more likely to engage in independent spending, while the formal party
committees should be less likely to spend. That is, we expect that well-resourced and highly
motivated donors faced with limits on contributions to the parties will find (or be offered) a
functionally equivalent way to give, thus displacing money that might otherwise have been
party money into non-party organizations affiliated with or allied with the parties.
8 We initially included the institutional powers of the governor (Beyle 2007, Ferguson 2012)
in our multivariate analysis. We expected that the higher the institutional powers of the governor, the more likely party and party network groups will engage in independent spending. However, the variable did not have a significant relationship with party and party network groups independent spending in gubernatorial races. Due to the small sample size, we excluded the gubernatorial institutional power variable from our final regression.
9 We used the National Conference on State Legislatures’ Redistricting Commissions:
Legislative Plans website to create this variable: http://www.ncsl.org/legislatures-elections/redist/2009-redistricting-commissions-table.aspx.
18
We separate our analysis for party committees and party-network groups (party-
affiliated groups and party-allied groups) in order to test our differing expectations for these
two types of groups regarding the contribution limits on donations to parties. Unfortunately,
we have only a limited number of cases on which to make an assessment. With this limitation
understood, we performed a regression analysis with the four previously discussed
independent variables.
[Table 2 goes here]
Three of the four variables are statistically significant. The level of candidate
contributions and the level of competition each relates significantly to spending in this arena,
with each showing the impact more strongly for party-affiliated and party-allied groups than
for the party organizations. Limiting contributions to parties does seem to depress
independent spending by party committees in gubernatorial elections, as expected, while at
the same time increasing independent spending by party-affiliated and party-allied groups.
However, we also need to remind the reader here that much of the money being spent by the
state parties in states without limits also came from the same national party-affiliated
organizations as are spending their money directly in the states with limits. Our final
independent variable, redistricting control, does not have a significant effect on either party
or party-network groups independent spending.
We can illustrate the interplay among these variables with two examples. Maine
meets both criteria for high party and party-network groups’ independent spending. In Maine,
the funds per vote were high in the gubernatorial election ($24.30 per vote) and the winning
candidate won with a small margin (1.76%). As expected, the level of independent spending
per vote by party and party-network groups in the Maine gubernatorial election was high
($16.90 per vote). In fact, party and party-network groups’ independent spending per vote in
19
Maine was highest of all of our states. As for the channels in which the spending took place:
Maine has no limits on donations to political party committee and party committees spent
$1.90 per vote in their independent spending. This was the highest level of independent
spending by formal party committees in any of gubernatorial elections in our states. Even so,
and despite the lack of a restriction on contributions to the parties, the party-affiliated and
party-allied groups outspent the formal party organizations by more than seven to one. The
“diversion” thesis about contribution limits just does not fit the facts of the case.
Idaho sits toward the other end of the scale. There was a fairly low level of candidate
funding ($9.52 per vote), and the victory margin was large (26.3%). Unsurprisingly,
independent spending by party and party-network groups in the Idaho gubernatorial election
was among the lowest ($0.27 per vote). The lack of any limits on donations to political
parties in Idaho suggests that such independent spending as there was should have been
channeled through the formal parties, and it was. No party-affiliated and party-allied groups
spent money independently in the election.
Clearly, candidate spending and competition influence the amount of independent
spending by party and party-network groups in gubernatorial races. The presence or absence
of contribution limits helps channel the paths that the money took, but the “diversion” thesis
does not fit the facts. Taken together, the findings also confirm that the party and party-
network groups tend to focus on competitive gubernatorial races generally, as opposed to
paying special attention to redistricting. Looking forward, it suggests that 2010 was not a
one-time spending spurt, but part of a longer pattern in which we should expect continued
growth.
20
Legislative Elections
Independent spending by party and party-network groups did not increase as much in
legislative as in gubernatorial elections.10 Even so, the level of spending was significant.
Therefore, we again use a multivariate analysis to examine spending by the party and party-
network groups in legislative elections across states.
Our dependent variables are the amounts of independent spending by party
committees and party-network groups (party-affiliated and party-allied groups) aggregated to
the chamber level. We again divide spending by total votes. The analysis excludes elections
in which the winning candidate received 75% or more of the vote.
As in the gubernatorial analysis, one independent variable is the candidates’ funding
per vote. Again, we expect higher candidate fundraising to be associated with higher
independent spending by party and party-network groups in legislative elections. We also
include control of the redistricting process to see whether party and party-network groups’
independent spending will be higher where the legislature has more control over redistricting.
The variable for political party contribution limits also remains the same. We expect
formal party committees to spend less and party-network organizations to spend more where
contributions to the formal party committees are limited.
The other independent variables differ from those in the gubernatorial analysis. First,
we include a variable indicating whether public financing with spending limits is available to
candidates. We expect that party committees and party-network groups will spend more in
chambers with public financing available to legislative candidates, since the candidates who
10
Party and party-network groups’ independent spending in gubernatorial elections more than doubled to $33.3 million in the 2010 election, while independent spending in legislative elections only increased by $2.1 million for a total of $20.7 million.
21
accept public funds must adhere to spending limits. Because of data limitations, our only
public funding states were Minnesota and Maine.
We also add a political variable measuring the percentage of seats the majority party
held in each chamber. We do so to include competition over control of the chamber in the
analysis. Our expectation is that independent spending by party and party-network groups
will be higher when the majority party holds a smaller percentage of seats. Spending should
increase when control of the chamber is more likely to be at stake (Moncrief 1992, Moncrief
and Patton 1993).
[Table 3 goes here.]
Table 3 displays the regression results for the party committee and party-network
groups’ independent spending in legislative elections. Unlike our results for the gubernatorial
elections, the legislative multivariate regressions do not clarify how the political and
institutional factors influence independent spending by party and party-network groups,
except with respect to public financing. The R2 in each legislative regression is much lower
than in the equivalent gubernatorial regression. None of the independent variables except one
significantly affects independent spending by party and party-network groups. The exception
is public financing, which does seem to be associated with higher levels of independent
spending by the formal party organizations. Interestingly, public financing does not
significantly affect independent spending by party-network groups. This may be because
both of the public funding states allow unlimited contributions to the parties.
It is surprising that neither the level of candidates’ receipts nor the size of the
majority party influences the party and party-network groups’ level of independent spending
in legislative elections. These results are perplexing because competitiveness, whether in a
race or in the chamber, routinely has been found to influence the level of spending in
22
elections. We plan to investigate this in future research. For now, we move on to explore
which party and party-network groups accounted for the increases in independent spending
from the 2006 to 2010 elections.
Party and Party-Network Groups in Four States
Independent spending by party and party-network groups in legislative elections did
not increase across all states. Of the fifteen states in our dataset, spending on legislative
elections by party and party-network groups actually went down in five. There were minor
increases in six others. In only four states did independent spending increase by more than
$750,000 – Maine, Michigan, North Carolina, and Tennessee. These are summarized below.
Maine: The formal state and legislative party committees (e.g., Maine Republican
Party, House Democratic Campaign Committee) were very active independent spenders in
Maine’s legislative elections. However, an almost $2 million increase in party-affiliated
independent spending accounted for most of the state’s large increase in independent
spending by party and party-network groups. The Republican State Leadership Committee
Maine PAC (an affiliate of the national RSLC) by itself was the source of an almost $2
million increase in spending in Senate races. Party-affiliated groups were even more involved
in Maine legislative elections than the independent spending data shows, because the Maine
Republican Party received about one-third of its funds in the 2010 election from the
Republican Governors Association (which is double what the RGA contributed to the state
party in 2006). Clearly, party-affiliated groups were key players in Maine’s elections.
North Carolina: The increase in party and party-network groups’ independent
spending in North Carolina was also due largely to the RSLC. The RSLC actually did not do
independent spending directly in North Carolina in 2010, but the group donated $1.25
23
million to Real Jobs NC, which we classified as party-affiliated for the election because it
was created in 2010 and received three-quarters of its money from the RSLC. Real Jobs NC
spent $1.5 million in independent expenditures in the 2010 election, making it the main
source of the increase in party-affiliated independent spending in that state.
Tennessee and Michigan: Another national Republican party-affiliated group, the
RGA, was the main source of the increases in independent spending in the Michigan and
Tennessee legislative elections. From the public records, it appears at first as if the state party
committees dramatically increased their independent spending. However, in both states, the
RGA was a major donor to the Republican Party committees. The Tennessee Legislative
Campaign Committee (TLCC) spent almost $900,000 on independent spending in the 2010
election. This large increase accounts for most of the increase in party and party-network
spending in the 2010 election. (The TLCC did not engage in independent spending at all in
the 2006 legislative election.) The increase in TLCC spending coincided with an influx of
money from the RGA: while only 3% of the TLCC’s money came from the RGA in 2006,
the RGA donated one-third of TLCC’s funds in 2010.
The situation in Michigan was similar. The Michigan Republican Party increased its
independent spending in legislative elections by $1.8 million from 2006 to 2010, making it
the most active party and party-network group in Michigan’s legislative elections. As in
Tennessee, this accounted for most of the increase in the state’s party and party-network
independent spending in 2010. But the state party was hardly acting alone. The RGA
donated 57% of the Michigan Republican Party’s funds in 2010 after donating nothing to the
state party in 2006. So, even though it might appear as if much of the increase was due to the
formal party committees in both states, one national party-affiliated organization (the RGA)
funded the increase.
24
This analysis makes it clear that a few national, party-affiliated groups were
responsible for the bulk of the increase in independent spending by both party and party-
network groups in the 2010 state elections. While a desire to control the redistricting process
was widely thought to have stimulated their activity in 2010, this did not show up as being
significant in the regressions. It is more likely that we are witnessing a change in the
organizational structure of American politics that is both durable and important. These
changes involve not only a deepening of networked relationships horizontally, within
jurisdictions, but a new, multi-jurisdictional form of vertical networking in which national
actors are deeply engaged in state elections across boundary lines.
Conclusions and Speculations
Unlike many of the statements that were initially made about the likely effects of
Citizens United, we find little evidence that businesses heavily increased their independent
spending in state elections between 2006 and 2010. Additionally, corporate and labor
independent spending increased mainly in states with no previous prohibitions on such
spending. Instead, we found that parties and party-network groups were the organizations
most actively engaged in independent spending before and after Citizens United. The lack
of a clear increase in business spending should not be a surprise. It is consistent with much
of what we have come to expect over the decades from corporate political behavior.
Much more significant were the findings about political parties and party-network
groups. Much of the writing about national political parties, state parties, non-party advocacy
organizations, and other interest groups, treats the organizations in a stylized way as if they
are clearly distinct. The world looks much more fluid to us. National organizations of state
party officials play a far greater role than they did a decade ago, to the point where they
25
dominated the growth in independent spending in the states over the four years covered here.
That is not to say the future will look like the past. Political actors in the business
sector may need more time to adapt to their opportunities after Citizens United. The data are
not yet available, at this writing, to analyze independent spending at the state level in the
elections of 2012. The early indications suggest, however, that the patterns of 2010 have
continued. More importantly, any adaptation by business and other actors would have to
occur within a system that is already undergoing a series of longer-term changes affecting the
roles of all party and non-party organizations. With 7,382 state legislative seats in the United
States (NCSL 2012), it is likely that the formal political party, party-affiliated and other
party-network organizations, with professional political staff and ongoing relationships
across the states, will maintain a crucial advantage over their organizational competitors.
These results may seem surprising or counterintuitive, but they highlight an important
development in American politics. Let us assume the early indications from 2012 are borne
out after full analysis, as we expect. In that case, the independent spending in the first two
sets of state elections after Citizens United would not be best understood as a manifestation
of corporate power flowing from a Supreme Court decision. Rather, the spending in these
elections seems to fit in more comfortably as a part of a longer, still developing story about
networked political parties. But the story, this time, has an important new wrinkle. The
parties are networked not only horizontally, across actors within jurisdictions. They are also
now networked vertically, with the key actors including national party organizations that
operate with fluidity across state elections.
26
REFERENCES
Beyle, Thad. 2007. “Gubernatorial Power: The Institutional Power Ratings for the 50
Governors of the United States”. http://www.unc.edu/~beyle/gubnewpwr.html.
Boatright, Robert G., Michael J. Malbin, Mark Rozell, Richard Skinner and Clyde Wilcox.
2003. “BCRA’s Impact on Interest Groups and Advocacy Organizations.” In Michael J.
Malbin, ed., Life after Reform: When the Bipartisan Campaign Reform Act Meets Politics.
Lanham MD: Rowman & Littlefield.
Boatright, Robert G., Michael J. Malbin, Mark Rozell, and Clyde Wilcox. 2006. “Interest
Groups and Advocacy Organizations after BCRA.” In Michael J. Malbin, ed. The Election
after Reform: Money, Politics, and the Bipartisan Campaign Reform Act. Lanham MD:
Rowman & Littlefield.
Clark-Muntean, Susan. 2008. “Corporate Contributions after the Bipartisan Campaign
Reform Act” Election Law Journal. 7(3): 233-244.
_____. 2011. “Corporate Independent Spending in the Post-BCRA to Pre-Citizens United
Era.” Business and Politics 13(1): Article 1.
Cohen, Marty, David Karol, Hans Noel and John Zaller. 2008. The Party Decides:
Presidential Nominations Before and After Reform. Chicago: University of Chicago Press
Confessore, Nicholas. 2014. “A National Strategy Funds State Political Monopolies.” New
York Times. January 11. Available at http://www.nytimes.com/2014/01/12/us/politics/a-
national-strategy-funds-state-political-monopolies.html. Accessed January 25, 2014.
Eismeier, Theodore J. and Philip H. Pollock III. 1988. Business Money and the Rise of
Corporate PACs in American Elections. New York: Quorum Books.
27
Ferguson, Margaret. 2012. “Governors and the Executive Branch.” in Politics in the
American States: A Comparative Analysis, eds. Virginia Gray, Russell L. Hanson, and
Thad Kousser. 10th ed. Washington, DC: CQ Press.
Issacharoff, Samuel and Jeremy Peterman. 2013. “Special Interests and Citizens United:
Access, Replacement, and Interest Group Response to Legal Change.” Annual Review of
Law and Social Science. 9:185-205.
Koger, Gregory, Seth Masket and Hans Noel. 2009. “Partisan Webs: Information Exchange
and Party Networks.” British Journal of Political Science. 39(3): 633-53.
La Raja, Raymond J. 2013a. “Why Super PACs: How the American Party System Outgrew
the Campaign Finance System.” The Forum. 10(4): 91-104.
_____. 2013b. “Richer Parties, Better Politics? Party-Centered Campaign Finance Laws and
American Democracy.” The Forum. 11(3): 313-338.
Masket, Seth. 2009. No Middle Ground: How Informal Party Organizations Control
Nominations and Polarize Legislatures. Ann Arbor MI: The University of Michigan
Press.
Michigan Campaign Finance Network. 2011. “The RGA’s $114M Shall Game”, March 15.
Available at http://www.mcfn.org/press.php?prId=130. Accessed August 15, 2012.
Moncrief, Gary F. 1992. “The Increase in Campaign Expenditures in State Legislative
Elections: A Comparison of Four Northwestern States.” The Western Political Quarterly.
45 (2): 549-558.
Moncrief, Gary F. and W. David Patton. 1993. “Upping the Campaign Ante as Parties
Compete to Control the State Legislature.” State & Local Government Review. 25 (1):
39-44.
28
Monroe, J.P. 2001. The Political Party Matrix: The Persistence of Organization. Ithaca:
State University of New York Press.
National Conference of State Legislatures (NCSL). 2009. “Redistricting Commissions:
Legislative Plans”. Available at http://www.ncsl.org/legislatures-elections/redist/2009-
redistricting-commissions-table.aspx. Accessed March 10, 2013.
________. 2012. “2012 States and Legislative Partisan Composition.” Available at
http://www.ncsl.org/documents/statevote/2012_Legis_and_State.pdf. Accessed August
23, 2012.
Skinner, Richard, Seth Masket and David Dulio. 2013. “527 Committees, Formal Parties
and Party Adaptation.” The Forum 11(2): 137-156.
Sorauf, Frank J. 1988. Money in American Elections. Scott, Foresman and Company.
Torres-Spelliscy, Ciara. 2012. "The $500 Million Question: Are the Democratic and
Republican Governors Associations Really State PACs under Buckley's Major Purpose
Test?" NYU Journal of Legislation & Public Policy 15: 485 (Spring).
Court cases:
Citizens United v. Federal Election Commission, 558 U.S. 50 (2010).
SpeechNow.org v. Federal Election Commission, 599 F. 3d (D.C. Cir., Mar. 26, 2010), cert.
denied 131 S. Ct. 553 (2010).
29
Table 1
Sources of Contributions to the RGA and DGA
Individual Corporate Labor Other
$
(Millions) %
$ (Millions)
% $
(Millions) %
$ (Millions)
%
RGA
2006 10.6 24% 30.8 73% 0.5 1% 0.8 2%
2010 44.2 38% 67.3 57% 0.5 0.40% 5.2 5%
DGA
2006 6 18% 19.5 57% 8 23% 0.5 1%
2010 8.2 15% 34.9 63% 11.6 21% 0.7 1%
Source: Analysis of reports filed with the Internal Revenue Service
30
Table 2
2010 Party and Party Network Groups’* Independent Spending in Gubernatorial Races**
Party Independent Spending Per Vote
Affiliated & Allied Groups’ Independent Spending Per Vote
Candidate Funds Per Vote 0.0349* 0.0837* (0.018) (0.045) Redistricting Control 0.082 0.302 (0.138) (0.343) Margin of Victory -0.0234* -0.0998** (0.013) (0.031)
Contribution Limits on Donations to Parties
-0.456* 1.154*
(0.247) (0.615) Constant 0.164 0.672 (0.405) (1.009) Observations 15 15
R-squared 0.550 0.694 Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1 *Party network groups include independent spending by party-affiliated groups and party-allies. ** CT & FL excluded because the target of IEs are not identified. MO, NC, & WA did not have gubernatorial elections in 2010.
31
Table 3
2010 Party and Party Network Groups* Independent Spending in Competitive Legislative Races**
Party Independent Spending Per
Vote
Affiliated & Allied Groups
Independent Spending Per Vote
Candidate Funds Per Vote -0.006 -0.015
(0.013) (0.013)
Redistricting Control 0.225 0.118 (0.201) (0.198)
Majority Party Share of Seats -0.019 -0.015 (0.016) (0.016)
Contribution Limits on Donations to Parties -0.326 0.487
(0.321) (0.316)
Public Financing with Spending Limit
1.085** -0.186
(0.484) (0.477) Constant 1.394 0.999
(1.133) (1.116)
Observations 34 34 R-squared 0.34 0.169 Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
*Party network groups include independent spending by party-affiliated groups and party-allies. **CT & FL excluded because the target of IEs are not identified. AZ also excluded because two-member districts complicate calculating race competitiveness.
32
Figure 1
37
61
107
145
-
20
40
60
80
100
120
140
160
Corp. Prohibited Corp. Prohibited Corp. Not
Prohibited
Corp. Not
Prohibited
2006 2010 2006 2010
Mil
lio
ns
Independent Spending 2006 v. 2010
State with/without Corporate Contribution
Prohibitions
33
Figure 2
Note: Crosshatched sections of the business bars shows spending in states where
business treasury independent spending was prohibited before Citizens United.
Crosshatched sections of the labor bars shows spending in states where labor
treasury independent spending was prohibited before Citizens United.
0
10
20
30
40
50
60
2006 2010 2006 2010 2006 2010 2006 2010 2006 2010 2006 2010
Party Party Affiliated Party Allied Ideology/Non-Party Business Labor
Mil
lio
ns
Independent Spending Across Sectors, 2006-
2010
34
Figure 3
-
5
10
15
20
25
30
Party Party
Affiliated
Party
Allied
Party Party
Affiliated
Party
Allied
Party Limited (n=4) Party Unlimited (n=16)
Mil
lio
ns
Independent Spending by Sector
Considering Party Contribution Limits by State
2006
35
Figure 4
0
10
20
30
40
50
60
70
Candidates Media/Consulting Operational
Expenses
Other Political
Groups
Party Committees
Mil
lio
ns
2010 RGA and DGA Contributions and
Expenditures
DGA
36
Appendix
Figure A-1
Figure A-2
-30
-25
-20
-15
-10
-5
0
5
10
15
OH TX MA TN AK MI AZ OK NC CO IA MN
x 1
00
00
0
Change in Business-Sector Independent Spending. 2006-2010
in States Prohibiting Corporate Spending in 2006
OH
-30
-20
-10
0
10
20
30
40
FL WA MO ME ID CA
x 1
00
00
0
Change in Business-Sector Independent Spending, 2006-2010
in States Permitting Corporate Spending in 2006
FL
